Inverted repeats (IRs) that can form a hairpin or cruciform structures are common in the human genome and may be sources of chromosome instability. Our previous studies revealed a paucity of closely spaced diverged Irs. We have been studying an IR involving the human Alu sequence (Alu-IR), the most abundant inverted repeat in the human genome, as a model of such structures in the yeast Saccharomyces cerevisiae. We also chose Alus because they are identified with many human diseases and are often seen at chromosomal breakpoints in cancer cells. We showed that a quasipalindrome comprised of two closely-spaced inverted Alu sequences is highly unstable in yeast. The Alu-IR caused a 1000-fold increase in homologous ectopic recombination between lys2 genes. We established that the ability of an Alu-IR to cause genome rearrangements results from a novel type of double-strand break in mitotic cells. Using a newly developed approach for analyzing rare mitotic chromosomal breaks, we demonstrated that resolution of a secondary structure formed by IRs results in broken molecules that are capped by a hairpin. We found that the requirements for repair of hairpin-capped DSBs were more extensive than for other DSBs, suggesting that they might be more threatening to the genome than other types of breaks. In addition to the standard DSB repair genes, repair of IR-induced breaks depended on the endonuclease activity of Mre11/Rad50/Xrs2 complex and on Sae2 protein. Failure of the mre11, rad50, xrs2 and sae2 mutants to open and process the hairpins led to the generation of chromosome inverted duplications. Based on these results we proposed a new role for the Mre11 complex in maintaining genome stability, namely in protecting the genome against hairpins. The role of Mre11 complex might be especially crucial in preserving the integrity of complex genomes, such as human, that contain a large amount of repetitive DNA capable of forming secondary structures. The system that we developed is now being expanded to address the impact of DNA double-strand breaks on the stimulation of changes at quasipalindromes. We are creating at DNA double-strand break target site next to inverted repeat pairs of Alus of varying levels of divergence. This system will enable us to investigate the consequences of a break in a region near large inverted repeats of the sort that are common in the human genome.